Seed Desiccation Sensitivity Varies with Geographic Distribution in Two New Zealand Native Pittosporum Species

Abstract

Ex situ conservation involves the maintenance and reproduction of species in areas outside their natural habitats. Seed banking is a well-established ex situ approach used for plant conservation. Seed banking consists of collecting, drying, and storing seeds to preserve genetic diversity. The main limitation of this technique is that the seed must be desiccation tolerant. Seed storage behaviour can vary among species, and for some species, drying without loss of viability is not possible, meaning storage under conventional seed banking conditions (sub-zero temperatures and low relative humidity) is not possible. Understanding seed storage behaviour is an essential prerequisite for establishing whether conventional seed banking is an option for seed conservation. This study investigated the desiccation tolerance and sensitivity of mature seeds of two native New Zealand species of Pittosporum (P. eugenioides and P. crassifolium) from two geographic locations (Palmerston North and Wellington), with the aim of understanding their seed storage behaviour and thereby improving conservation outcomes. The variables measured were seed moisture content, viability, germination, and desiccation responses. We developed sorption isotherm for both species to support the future development of storage protocols. Our results show that both P. eugenioides and P. crassifolium display non-orthodox behaviour, i.e., are desiccation sensitive and cannot be stored under conventional seed banking conditions, but also suggest that seed desiccation responses vary with the geographical origin of the seeds. This study highlights the importance of exploring seed storage behaviour using different populations to optimize ex situ conservation strategies aimed at preserving the genetic diversity of New Zealand’s threatened and endangered species.

Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.